Comparison of clinical parameters, laboratory parameters, GPC4 levels, and CLU levels between women with PCOS and healthy women
The clinical parameters, laboratory parameters, GPC4 levels, and CLU levels are summarized in Table 1. The distributions of GPC4 and CLU concentrations in PCOS and healthy women are shown in Figures 1a and 1b. Serum GPC4 and CLU levels in PCOS patients were significantly higher than those in healthy women (both P < 0.001).
Comparison of the relative risk of PCOS at different serum GPC4 and CLU levels
According to the level of GPC4, all subjects were divided into the following three groups from low to high: T1 group (GPC4<1.29 ng/ml), T2 group (1.29≤ GPC4<1.80 ng/ml), and T3 group (GPC4 ≥1.80 ng/ml). The results showed that the risks of PCOS in the T3 and T2 groups were significantly higher than that in the T1 group (OR =12.6, 95% CI: 3.443-46.107, P<0.001; OR= 4.9, 95% CI: 1. 413-16.988, P=0.012), but there was no significant difference in the risk of PCOS between the T3 and T2 groups (OR= 2.571, 95% CI: 0.813-8.134, P=0.108) (Figure 2a).
According to the level of CLU, all subjects were divided into the following three groups from low to high: Q1 group (CLU<257.65 µg/ml), Q2 group (257.65≤ CLU<442.27 µg/ml), and Q3 group (CLU ≥442.27 µg/ml). The results showed that the risks of PCOS in the Q3 and Q2 groups were significantly higher than that in the Q1 group (OR= 625, 95% CI: 37.005-10556.023, P<0.001; OR= 25, 95% CI: 2.966~210.717, P=0.003), and the risk of PCOS in the Q3 group was significantly higher than in the Q2 group (OR= 25, 95% CI: 2.966-210.717, P=0.003) (Figure 2b).
Table 1 Clinical parameters, laboratory parameters, GPC4 levels, and CLU levels in PCOS and healthy women
Variables
|
Controls(n=40)
|
PCOS (n=40)
|
P
|
年龄
|
27.5±3.63
|
26.2±4.69
|
0.170
|
BMI(kg/㎡)
|
21.40±1.76
|
24.03±3.64
|
0.001*
|
FSH(mIU/ml)
|
5.42±1.39
|
5.27±1.70
|
0.672
|
LH(mIU/ml)
|
4.95(3.88-6.18)
|
6.15(4.38-9.18)
|
0.012*
|
LH/FSH
|
0.95(0.72-1.13)
|
1.22(1.01-1.51)
|
<0.001*
|
E2(pg/ml)
|
51.52(48.34-55.55)
|
45.71(35.13-62.70)
|
0.068
|
TT(ng/dl)
|
25.01(22.10-27.76)
|
38.17(31.73-51.10)
|
<0.001*
|
FPG(mmol/l)
|
4.84(4.46-5.08)
|
5.73(5.17-6.44)
|
<0.001*
|
FINS(uU/ml)
|
8.68(7.37-9.03)
|
19.70(12.97-29.93)
|
<0.001*
|
HOMA-IR
|
1.82(1.50-2.00)
|
4.75(3.57-8.69)
|
<0.001*
|
TC(mmol/l)
|
3.86(3.55-4.28)
|
4.39(4.06-5.24)
|
<0.001*
|
TG(mmol/l)
|
0.82(0.65-1.26)
|
1.53(1.19-2.20)
|
<0.001*
|
HDL-C(mmol/l)
|
1.46±0.27
|
1.17±0.26
|
<0.001*
|
LDL-C(mmol/l)
|
2.21(2.06-2.54)
|
2.80(2.52-3.44)
|
<0.001*
|
GPC4(ng/ml)
|
1.30±0.61
|
1.82±0.49
|
<0.001*
|
CLU(ug/ml)
|
228.59±82.42
|
468.79±92.85
|
<0.001*
|
A p value of <0.05 was considered significant (*).
Correlation analysis of serum GPC4 and CLU with general data and clinical indicators
We next performed a correlation analysis of serum GPC4, serum CLU and other variables. In the control group, serum GPC4 was positively correlated with BMI, E2, FINS, and HOMA-IR (P<0.05), and serum CLU was positively correlated with BMI, E2, FPG, and HOMA-IR (P<0.05). In the PCOS group, serum GPC4 was positively correlated with FSH, FPG, FINS, HOMA-IR, and TG (P<0.05), and whereas serum CLU was positively correlated with BMI, FPG, FINS, and HOMA-IR (P<0.05) (Table 2).
The natural logarithmic conversion of HOMA-IR conformed to a normal distribution, and the scatterplot indicated that LnHOMA-IR was significantly linearly correlated with serum GPC4 (r=0.74, P<0.001) and CLU (r=0.589, P<0.001) in PCOS patients (Figure 3a and 3b).
Table 2. Correlation analysis of serum GPC4 and CLU with general data and clinical indicators
Variables
|
GPC4
|
CLU
|
Controls
|
PCOS
|
Controls
|
PCOS
|
r
|
P
|
r
|
P
|
r
|
P
|
r
|
P
|
年龄
|
-0.024
|
0.885
|
-0.241
|
0.134
|
-0.158
|
0.329
|
0.075
|
0.648
|
BMI(kg/㎡)
|
0.616
|
<0.001*
|
0.304
|
0.056
|
0.474
|
0.002*
|
0.552
|
<0.001*
|
FSH(mIU/ml)
|
0.092
|
0.572
|
0.338
|
0.033*
|
0.296
|
0.064
|
0.261
|
0.104
|
LH(mIU/ml)
|
-0.052
|
0.750
|
0.123
|
0.449
|
0.271
|
0.091
|
0.258
|
0.108
|
LH/FSH
|
-0.159
|
0.327
|
-0.176
|
0.277
|
-0.041
|
0.802
|
-0.038
|
0.818
|
E2(pg/ml)
|
0.365
|
0.020*
|
-0.072
|
0.659
|
0.423
|
0.006*
|
-0.097
|
0.550
|
TT(ng/dl)
|
0.244
|
0.128
|
0.139
|
0.393
|
0.051
|
0.753
|
0.201
|
0.213
|
FPG(mmol/l)
|
0.254
|
0.114
|
0.384
|
0.014*
|
0.373
|
0.018*
|
0.363
|
0.021*
|
FINS(uU/ml)
|
0.325
|
0.041*
|
0.729
|
<0.001*
|
0.281
|
0.078
|
0.639
|
<0.001*
|
HOMA-IR
|
0.360
|
0.023*
|
0.706
|
<0.001*
|
0.384
|
0.015*
|
0.639
|
<0.001*
|
TC(mmol/l)
|
0.149
|
0.359
|
0.059
|
0.716
|
0.010
|
0.950
|
0.222
|
0.169
|
TG(mmol/l)
|
0.106
|
0.514
|
0.372
|
0.018*
|
0.165
|
0.310
|
0.211
|
0.191
|
HDL-C(mmol/l)
|
-0.115
|
0.481
|
-0.247
|
0.124
|
-0.307
|
0.054
|
-0.078
|
0.635
|
LDL-C(mmol/l)
|
0.204
|
0.206
|
0.138
|
0.396
|
0.191
|
0.238
|
0.295
|
0.065
|
CLU(ug/ml)
|
0.493
|
0.001*
|
0.493
|
0.001*
|
-
|
-
|
-
|
-
|
* The difference was statistically significant at P<0.05 or P<0.001.
Multiple linear regression analysis of serum GPC4, serum CLU, and related factors in patients with PCOS
GPC4 was considered the dependent variable, and meaningful variables in the correlation analysis were included. Because there was a functional relationship among FINS, FPG, and HOMA-IR, only HOMA-IR, FSH, and TG were included as independent variables for multiple stepwise linear regression analysis. The results showed that HOMA-IR was an independent correlation factor of serum GPC4 in PCOS patients (P<0.001) (Table 3).
CLU was used as the dependent variable, and meaningful variables in the correlation analysis were included because FINS, FPG, and HOMA-IR were functionally related. Only BMI and HOMA-IR were included as independent variables, and multiple stepwise linear regression analysis showed that BMI and HOMA-IR were independent correlated factors in serum CLU in PCOS patients (P<0.05 or P<0.001) (Table 4).
Table 3 Multiple linear regression analysis of serum GPC4 and related factors in PCOS patients
Variables
|
B
|
SE
|
β
|
t
|
P
|
95%CI
|
Constant
|
1.367
|
0.093
|
-
|
14.698
|
<0.001*
|
1.179-1.556
|
HOMA-IR
|
0.067
|
0.011
|
0.699
|
6.028
|
<0.001*
|
0.044-0.089
|
*p value less than 0.05 was considered statistically significant.
Table 4 Multiple linear regression analysis of serum CLU and related factors in PCOS patients
Variables
|
B
|
SE
|
β
|
t
|
P
|
95%CI
|
Constant
|
257.575
|
66.460
|
-
|
3.876
|
<0.001*
|
122.788-392.361
|
HOMA-IR
|
7.774
|
2.082
|
0.500
|
3.733
|
0.001*
|
3.551-11.998
|
BMI
|
6.873
|
2.948
|
0.312
|
2.331
|
0.025*
|
0.893-12.853
|
*p value less than 0.05 was considered statistically significant.
GPC4 is an adipokine in the proteoglycan family that binds directly to insulin receptors and acts as an insulin receptor sensitizer to enhance insulin signaling; when GPC4 is depleted, it weakens insulin receptor phosphorylation and downstream signaling[21]. The present study found that compared to healthy controls, PCOS patients had elevated serum GPC4 levels. The study subjects were divided into three groups according to the concentration of GPC4 by the third-percentile method. We found that the risk of PCOS in both the T3 and T2 groups was significantly higher than that in the T1 group, but there was no statistically significant difference in the risk of PCOS in the T2 group and T3 group, indicating that the risk of PCOS development may be related to serum GPC4 levels. Correlation analysis showed that in the control group, serum GPC4 was positively correlated with BMI, E2, FINS, HOMA-IR, and CLU, while in the PCOS group, GPC4 was positively correlated with FSH, FPG, FINS, HOMA-IR, TG, and CLU. Further linear regression analysis found that HOMA-IR was an independent correlation factor of serum GPC4 in PCOS patients, indicating that serum GPC4 may be a marker related to insulin resistance in PCOS patients. The results of the present study were similar to those of previous studies. Jędrzejuk et al.[22] found that serum GPC4 in PCOS patients is higher than that in healthy controls and is associated with cardiovascular disease risk factors, such as BMI, waist circumference, WHR, fasting insulin, HOMA-IR, and fat distribution. Similarly, Altinkaya et al.[23]. studied the serum GPC4 level of PCOS patients and found that the serum GPC4 of women in the PCOS group is significantly higher than that of the control group and is positively correlated with BMI, FPG, HOMA-IR, TG and TT, and they reported that BMI is an independent influencing factor of GPC4, which may be due to population variability and small sample size. At present, the mechanism of action of GPC4 in metabolic regulation is not fully understood. Ussar S et al. [21] speculated that the increase in GPC4 levels in pathological conditions may be a new regulatory mechanism through which fat counteracts insulin resistance, and maintaining high levels of serum GPC4 in patients with severe insulin resistance or diabetes may reduce insulin requirements. At the same time, it has been proposed that GPC4 may also affect adipocyte differentiation and enhance insulin receptor signaling through Wnt signaling. However, there are still relatively few studies on GPC4 and PCOS, and their mechanism of action is still not completely clear. More basic and prospective studies are needed to explore the relationship between the two in the future.
CLU, also known as APOJ, is a heterodimer sulfated glycoprotein, and it is also considered an adipocytokine. CLU mainly comprises two forms as follows: 1) oxidative stress-induced nonglycosylated nucleonated cell homologous 55 kDa variant (nCLU), consisting of parallel α and β chains; and 2) secretory or cytoplasmic variant (sCLU), which is cleaved by proteolysis, linked by five disulfide bonds, and released from the cell in an antiparallel manner[24]. sCLU is primarily cycled as a component of HDL-C, and in healthy subjects, higher sCLU levels are associated with cardioprotective HDL-C, suggesting that sCLU may play a role in preventing vascular disease progression[25]. While nCLU primarily promotes ionizing radiation-induced cell death and triggers apoptosis in a BAX-dependent mechanism, it has not been found to be associated with cardiometabolic pathology[26]. CLU is involved in a wide range of pathological and physiological processes, including cellular senescence, tumorigenesis, tumor development, complement regulation, sperm maturation, and lipid transport[27]. In the present study, the serum CLU level of PCOS patients was significantly higher than that of healthy women, and the study subjects were grouped into tertiles according to the concentration of serum CLU. It was found that the relative risk of PCOS gradually increased with the increase in tertiles of CLU concentration, indicating that the serum CLU level may be closely related to the relative risk of PCOS. In the control group, serum CLU was positively correlated with BMI, E2, FPG, HOMA-IR, and GPC4, while in women in the PCOS group, serum CLU was positively correlated with BMI, FPG, FINS, HOMA-IR, and GPC4. Stepwise linear regression analysis found that BMI and HOMA-IR were independent correlated factors in PCOS patients. The results of the present study were similar to those of Seo J A et al.[28], who found that serum CLU levels are elevated in 20 women with PCOS and positively correlated with FINS, free fatty acids, and HOMA-IR, suggesting that CLU is a liver factor that regulates muscle glucose metabolism and insulin sensitivity.
It is currently believed that the mechanism of action of serum CLU involved in metabolic diseases may be through several pathways. First, CLU may be an anorexia factor, which is a key point for the hypothalamus to control energy metabolism. Animal studies have found that administration of CLU to mouse centers leads to anorexia, weight loss, and activation of signal transduction-activated transcript-3 (STAT3) in mice. Conversely, suppression of hypothalamic CLU increases food intake and body weight, which may lead to obesity. This mechanism of action may increase the sensitivity of hypothalamic leptin receptors through CLU binding to low-density lipoprotein receptor-related protein-2 (LRP2), which in turn promotes leptin signaling, resulting in reduced energy intake and increased consumption[29]. Second, the expression of fat cell genes and proteins of CLU in obese patients is upregulated, and CLU inhibits insulin signaling after binding to LRP2 receptors in the liver, which promotes hepatic gluconeogenesis through upregulation of GCK and PKLR as well as downregulation of SREBP1, thereby reducing apolipoprotein A1 expression and decreasing HDL-C, ultimately promoting the occurrence of fatty liver[30]. In addition, CLU also acts as a chemoattractant for the directed migration of macrophages, stimulating the expression and secretion of chemotactic cytokines, such as TNF-α, allowing CLU to act as a bridge between inflammation and tissue remodeling by recruiting immune cells[31]. The present study found a significant independent positive correlation of serum CLU with HOMA-IR and BMI in PCOS patients, suggesting that CLU may play a key role in energy metabolism and insulin resistance. However, there are still few studies on CLU and PCOS. The specific mechanism is still unclear, and the relationship between serum CLU and PCOS as well as its mechanism of action need to be further explored.
Serum GPC4 and CLU are important adipocyte cytokines, and the present study found that there is abnormal expression of both in PCOS patients and that there is a certain correlation between the two, both of which are independently positively correlated with HOMA-IR. Thus, these findings suggested that GPC4 and CLU may jointly affect insulin sensitivity through different insulin signaling pathways, indicating that they may be markers of insulin resistance in PCOS patients.